Advertisement reviewing method, device and equipment and storage medium

Abstract

The application discloses an advertisement auditing method and device, equipment and a storage medium, the method comprises the following steps: generating a smart contract based on a preset advertisement auditing rule and a preset listing rule, and publishing the smart contract to a block chain; performing automatic auditing based on advertisement material data and an advertiser data, and obtaining an automatic auditing result; obtaining an auditing index and sorting advertisement auditors to obtain an advertisement auditor data list; determining a target advertisement auditor according to the advertisement auditor data list, distributing the advertisement material to the target advertisement auditor for manual auditing through the smart contract, and obtaining a manual auditing result; publishing the automatic auditing result and the manual auditing result to the block chain, and calculating the advertisement auditing result through the smart contract. The problems of low advertisement auditing efficiency and the risk of human interest exchange are solved, the auditing efficiency is improved, the interest exchange between the auditors and the advertisers is avoided, and the fair and just listing and ranking of the compliant advertisements are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of advertisement review, and in particular to an advertisement review method, apparatus, device and storage medium. Background Art

[0002] Currently, ad review primarily involves AI-based and manual review. AI review is currently limited in its ability to review H5 ad pages and video ads, requiring manual review. While proposed blockchain-based ad data processing methods utilize blockchain-based methods for ad review, these methods merely store the results of a single manual review on a blockchain platform. This process still suffers from inefficiencies and the artificial exchange of interests between reviewers and advertisers, which can negatively impact ad placement and ranking. Summary of the Invention

[0003] The main purpose of the present invention is to provide an advertising review method, device, equipment and storage medium, aiming to solve the technical problems of low efficiency of advertising review in the prior art and the existence of space and risks for artificial exchange of interests between reviewers and advertisers.

[0004] To achieve the above object, the present invention provides an advertisement review method, which comprises the following steps:

[0005] Generate a smart contract based on preset advertising review rules and preset listing rules, and publish the smart contract to the blockchain;

[0006] When receiving the advertising material data and the advertiser data sent by the advertiser, performing an automated review based on the advertising material data and the advertiser data to obtain an automated review result;

[0007] Obtain the review index of each advertising reviewer and sort the advertising reviewers according to the review index to obtain a list of advertising reviewer data, wherein the review index is related to the degree to which the smart contract recommends the corresponding advertising reviewer to conduct advertising review;

[0008] Determine a target advertising reviewer based on the advertising reviewer data list, and assign the advertising material to the target advertising reviewer for manual review through the smart contract to obtain a manual review result;

[0009] The automated review result and the manual review result are published to the blockchain, and the smart contract is used to calculate based on the automated review result and the manual review result to obtain the advertisement review result.

[0010] Optionally, the step of obtaining the audit index of each advertising auditor and sorting the advertising auditors according to the audit index to obtain an advertising auditor data list includes:

[0011] obtaining advertisement auditor list data, wherein the advertisement auditor list data comprises an advertisement auditor list and advertisement types that can be audited by each advertisement auditor;

[0012] screening the advertisement auditor list according to the advertisement types and the advertisement types that can be audited by each advertisement auditor, to obtain a target advertisement auditor list;

[0013] obtaining an audit index of each advertisement auditor in the target advertisement auditor list;

[0014] ranking the advertisement auditors in the target advertisement auditor list according to the audit indexes of the advertisement auditors, to obtain an advertisement auditor data list.

[0015] Optionally, the target advertisement auditor is determined according to the advertisement auditor data list, and advertisement materials are distributed to the target advertisement auditor for manual auditing through the smart contract, to obtain a manual auditing result, comprising:

[0016] obtaining a length of the advertisement auditor data list and a number threshold of advertisement materials sent at a time;

[0017] comparing the length of the advertisement auditor data list with the number threshold of advertisement materials sent at a time;

[0018] if the length of the advertisement auditor data list is greater than the number threshold of advertisement materials sent at a time, determining an audit threshold according to the audit indexes of the advertisement auditors in the advertisement auditor data list;

[0019] grouping the advertisement auditors in the advertisement auditor data list according to the audit threshold, to obtain a first group and a second group, wherein the audit indexes of the advertisement auditors in the first group are greater than the audit indexes of the advertisement auditors in the second group;

[0020] calculating the numbers of the advertisement auditors in the first group and the second group, to obtain a first number and a second number, wherein the first number is greater than the second number;

[0021] screening the advertisement auditors in the first group and the second group according to the first number and the second number, to obtain a target advertisement auditor;

[0022] distributing advertisement materials to the target advertisement auditor for manual auditing through the smart contract, to obtain a manual auditing result.

[0023] Optionally, the calculating the number of advertisement auditors in the first group and the second group to obtain a first number and a second number comprises:

[0024] disassembling the number threshold to obtain a plurality of calculation number pairs;

[0025] randomly sorting the plurality of calculation number pairs to obtain a calculation number pair sequence;

[0026] determining a first random number according to the number threshold, wherein the first random number is less than the number threshold;

[0027] obtaining a target calculation number pair from the calculation number pair sequence according to the first random number;

[0028] determining a first number and a second number according to the target calculation number pair.

[0029] Optionally, before the obtaining the audit index of each advertisement auditor, the method further comprises:

[0030] when detecting that a user uploads feedback data after watching an advertisement, obtaining an advertisement audit result of the advertisement watched by the user and an initial audit index of a corresponding advertisement auditor;

[0031] calculating, according to the feedback data after watching the advertisement, the advertisement audit result of the advertisement watched by the user and the initial audit index of the corresponding advertisement auditor, to obtain a current audit index of the corresponding advertisement auditor;

[0032] updating the initial audit index according to the current audit index to obtain the audit index of each advertisement auditor.

[0033] Optionally, the obtaining an automatic audit result based on the advertisement material data and the advertiser data comprises:

[0034] determining an advertisement type according to the advertisement material data;

[0035] performing compliance detection on the advertisement material according to the advertisement type to obtain a detection result, wherein the compliance detection at least comprises at least one of text compliance detection, audio compliance detection, image / video compliance detection and H5 page security compliance detection;

[0036] performing automatic audit based on the advertisement material data, the advertiser data and the detection result to obtain an automatic audit result.

[0037] Optionally, after the publishing the automatic audit result and the manual audit result to the blockchain and performing calculation according to the automatic audit result and the manual audit result through the smart contract to obtain an advertisement audit result, the method further comprises:

[0038] when the advertising review result is passed, changing a state of the advertising material to a reviewed state through the smart contract;

[0039] detecting the advertising principal data, and determining whether the advertising principal has a fee default behavior according to the advertising principal data;

[0040] if not, determining whether the advertising principal has a historical advertising rule violation rectification behavior according to the advertising principal data;

[0041] if not, automatically uploading the advertising material through the smart contract.

[0042] In addition, to achieve the above object, the present application also provides an advertising review device, which comprises:

[0043] a publishing module, configured to generate a smart contract based on preset advertising review rules and preset uploading rules, and publish the smart contract to a block chain;

[0044] an auditing module, configured to, when receiving advertising material data and advertising principal data sent by an advertising principal, perform automatic auditing based on the advertising material data and the advertising principal data, and obtain an automatic auditing result;

[0045] a sorting module, configured to obtain an auditing index of each advertising reviewer, and sort the advertising reviewers according to the auditing index, and obtain an advertising reviewer data list, wherein the auditing index is related to a recommendation degree of the smart contract in recommending corresponding advertising reviewers to perform advertising review;

[0046] the auditing module is further configured to determine a target advertising reviewer according to the advertising reviewer data list, and distribute the advertising material to the target advertising reviewer for manual auditing through the smart contract, and obtain a manual auditing result;

[0047] a determining module, configured to publish the automatic auditing result and the manual auditing result to the block chain, and calculate the advertising review result according to the automatic auditing result and the manual auditing result through the smart contract.

[0048] In addition, to achieve the above object, the present application also provides an advertising review device, which comprises: a memory, a processor, and an advertising review program stored in the memory and capable of running on the processor, wherein the advertising review program is configured to implement the steps of the advertising review method as described above.

[0049] In addition, to achieve the above-mentioned purpose, the present invention further proposes a storage medium, on which an advertisement review program is stored. When the advertisement review program is executed by a processor, the steps of the advertisement review method described above are implemented.

[0050] The present invention generates a smart contract based on preset advertising review rules and preset listing rules, and publishes the smart contract to a blockchain; when receiving advertising material data and advertiser data sent by an advertiser, performs automated review based on the advertising material data and the advertiser data to obtain an automated review result; obtains a review index of each advertising reviewer, and sorts the advertising reviewers according to the review index to obtain an advertising reviewer data list, wherein the review index is related to the degree of recommendation of the corresponding advertising reviewer for advertising review by the smart contract; determines a target advertising reviewer according to the advertising reviewer data list, and assigns the advertising material to the target advertising reviewer for manual review through the smart contract to obtain a manual review result; publishes the automated review result and the manual review result to the blockchain, and calculates the automated review result and the manual review result through the smart contract to obtain an advertising review result. Through the above method, the smart contract is published to the blockchain, and the advertising materials are subject to automatic review and manual review respectively through the smart contract. The smart contract is used to calculate according to the automatic review results and the manual review results to obtain the advertising review results. This solves the problems of low advertising review efficiency and the existence of space and risks for artificial exchange of interests between reviewers and advertisers. It improves the review efficiency, avoids the exchange of interests between reviewers and advertisers, and ensures the fair and just listing and ranking of compliant advertisements. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a structural diagram of an advertisement review device in a hardware operating environment according to an embodiment of the present invention;

[0052] Figure 2 This is a flow chart of a first embodiment of the advertisement review method of the present invention;

[0053] Figure 3 This is a flowchart of automatic advertisement review and listing in an embodiment of the advertisement review method of the present invention;

[0054] Figure 4 This is a design diagram of the smart contract data structure in an embodiment of the advertising review method of the present invention;

[0055] Figure 5 This is a flow chart of a second embodiment of the advertisement review method of the present invention;

[0056] Figure 6 This is a diagram of the underlying service deployment architecture of the consortium chain blockchain in an embodiment of the advertising review method of the present invention;

[0057] Figure 7 This is a structural block diagram of the first embodiment of the advertisement review device of the present invention.

[0058] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0059] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0060] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of an advertisement review device in the hardware operating environment involved in an embodiment of the present invention.

[0061] like Figure 1 As shown, the advertising review device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to implement connection and communication between these components. The user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard). Optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wireless-Fidelity (Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk storage. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0062] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation to the advertising review device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0063] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module and an advertisement review program.

[0064] exist Figure 1In the advertising review device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the advertising review device of the present invention can be set in the advertising review device, and the advertising review device calls the advertising review program stored in the memory 1005 through the processor 1001, and executes the advertising review method provided by the embodiment of the present invention.

[0065] The embodiment of the present invention provides an advertisement review method, referring to Figure 2 , Figure 2 Schematic diagram of the flow of the first embodiment of the advertisement review method of the present invention.

[0066] In this embodiment, the advertisement review method includes the following steps:

[0067] Step S10: Generate a smart contract based on preset advertising review rules and preset listing rules, and publish the smart contract to the blockchain.

[0068] It should be noted that the execution subject of this embodiment is an advertisement review device, and may also be other devices with the same or similar functions. This embodiment does not limit this. This embodiment is described by taking the advertisement review device as an example.

[0069] It can be understood that in this embodiment, the advertising review method can be applied to an advertising review device. In the advertising review device, three types of roles are involved, namely advertisers, operators, blockchain platforms and AI platforms. Among them, advertisers and operators are real people, and the blockchain platform is the underlying support service system. Operators do not need to review advertisements and put advertisements on the operation platform. Operators only need to check the review results of each advertisement on the operation platform when needed. Advertisers can complete at least five types of tasks in the advertising configuration platform, namely login, advertising configuration and submission, viewing of advertising review results, viewing of advertising revenue, and payment. The blockchain platform and AI platform are responsible for automated review and notification of review results.

[0070] In the specific implementation, the advertising review rules and listing rules are encapsulated into smart contracts and then packaged into blocks, which are published to the blockchain through a node in the alliance chain; the alliance chain nodes of the blockchain complete the consensus of the smart contract and store the smart contract locally on each node.

[0071] It's worth noting that blockchain is a new application model for computer technologies, including distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Also known as distributed ledger technology, it's an internet database technology characterized by decentralization, openness, and transparency, allowing everyone to participate in database records. Consortium blockchains are limited to members of a specific group and a limited number of third parties. They designate multiple pre-selected nodes as bookkeepers, and the generation of each block is determined by all pre-selected nodes. Other connected nodes can participate in transactions but are not subject to the bookkeeping process. Other third parties can conduct limited queries through the blockchain's open API. A smart contract is a set of digitally defined promises, including the protocols by which the contract participants can execute these promises. These contracts are visible to all users on the blockchain. In other words, smart contracts are based on trusted, immutable data and are based on pre-defined rules and terms that can be automatically executed.

[0072] Step S20: When the advertising material data and the advertiser data sent by the advertiser are received, an automated review is performed based on the advertising material data and the advertiser data to obtain an automated review result.

[0073] It should be noted that the advertiser logs in to the advertising configuration platform. If the login is successful, the advertiser will complete the advertising configuration on the advertising configuration platform, click the Submit button to submit the advertising material, and upload the advertising material and advertiser data (such as advertiser number, payment status, etc.).

[0074] like Figure 3 As shown, Figure 3This is a flowchart of automatic review and listing of advertisements in the advertising review method of this embodiment. In the automatic review and listing process of advertisements, three types of roles are involved, namely advertisers, operators, blockchain platforms and AI platforms. Among them, advertisers and operators are real people, and the blockchain platform is the underlying support service system. Operators do not need to review advertisements and list advertisements on the operation platform. Operators only need to check the review results of each advertisement on the operation platform when necessary. Advertisers can complete at least five types of tasks in the advertising configuration platform, namely login, advertisement configuration and submission, viewing of advertisement review results, viewing of advertisement revenue, and payment. The blockchain platform and AI platform are responsible for automated review and notification of review results. Advertisement review and listing rules are encapsulated into smart contracts and then packaged into blocks, which are published to the blockchain through a node of the alliance chain. The blockchain alliance chain nodes complete the consensus of the smart contract and store the smart contract locally on each node. The advertiser logs in to the ad configuration platform. If the login is successful, the advertiser completes the ad configuration on the ad configuration platform and submits the ad application. The materials submitted by the advertiser together with the advertiser's relevant information (such as advertiser number, payment status, etc.) are distributed to the AI ​​detection module through the message queue MQ such as Kafka. The AI ​​detection module detects the ad materials according to the detection rules and determines whether the detection is successful or not. If successful, for ads with weak AI recognition capabilities such as H5 page ads and video ads, the relevant ad materials are sent to multiple people for manual review through smart contracts. After the manual review results are sent back to the blockchain, the ad materials, advertiser information and AI detection The module detection results and multi-person review results are sent to the blockchain platform for packaging and block packaging, and are published to the blockchain by a node of a certain blockchain. The nodes of the alliance chain reach consensus and save the advertising materials, advertiser information and AI detection module detection results. The smart contract is triggered, and the advertising materials, advertiser information and AI detection module detection results are used as input parameters to execute the smart contract; the smart contract will automatically calculate whether the advertising materials can pass the review through rules. If they can pass the review, the listing status of the advertisement will be marked as on the shelf; the execution results (including whether the review results pass or not and whether they are on the shelf) are published to the auditors and advertisers through the message queue Kafka. The auditors can log in to the advertising operation platform to view the review results and listing status of AI and smart contracts; advertisers can log in to the advertising configuration platform to view whether their recently uploaded advertisements have passed the review, and can also view the revenue data of currently listed advertisements.

[0075] Furthermore, the performing of automated review based on the advertising material data and the advertiser data to obtain an automated review result includes: determining an advertisement type based on the advertising material data;

[0076] Perform compliance testing on the advertising material according to the advertising type to obtain a test result, wherein the compliance testing includes at least one of text compliance testing, audio compliance testing, image and video compliance testing, and H5 page security compliance testing; perform automated review based on the advertising material data, the advertiser data, and the test result to obtain an automated review result.

[0077] It should be noted that when the advertising material data and advertiser data sent by the advertiser are received, the advertising material is detected according to the detection rules through the AI ​​detection module, that is, AI detection of the advertisement is performed first, rather than calling the AI ​​module detection separately in each alliance chain node in the blockchain. By reducing repeated detection, the number and cost of AI detection are reduced.

[0078] As shown in Table 1, Table 1 is the audit rules table for AI and smart contracts, which includes modules, content items and main methods. The modules include AI and smart contracts. The detection content of AI includes text compliance detection, audio compliance detection, image and video compliance detection and H5 page security compliance detection. The detection content of smart contracts includes advertiser status detection, advertising material detection, advertiser payment detection, and historical advertising compliance detection. The main method of text compliance detection includes OCR recognition of text in pictures, videos, and H5 pages, sensitive word detection, and blacklist word matching; the main method of audio compliance detection includes converting speech into text, and performing sensitive word detection and blacklist word matching; the main method of image and video compliance detection includes detecting whether the image has sensitive content through image recognition; the main method of H5 page security detection includes detecting web pages such as XSS attacks. <script>标签和代码带来的攻击行为；广告主状态检测的主要方式包括检测广告主是否被封禁，广告素材检测的主要方式包括检测素材参数的完备性，广告主缴费检测的主要方式包括检查广告主是否欠费，历史广告合规检测的主要方式包括检查广告主是否有整改中的广告。

[0079] 表1

[0080]

[0081] 值得说明的是，AI检测的检测结果分别为文字合规检测数据、音频合规检测数据、图像视频合规检测数据、H5页面安全合规检测数据。其中，文字合规检测数据主要将通过OCR技术将图片、视频、H5网页中的文字信息识别和提取出来，对文字分词后，对比敏感词库，进行识别，若识别结果不涉及敏感词，则ad_word设置为0，否则设置为1；音频合规检测数据主要是通过语音转文字技术将语音识别为文字，对文字分词后，对比敏感词库，进行识别，若识别结果不涉及敏感词，则ad_sound设置为0，否则设置为1；图像视频合规检测数据主要是通过图像识别技术，对比内容库，如果识别结果不包含违规内容，则ad_photo(设置为0，否则设置为1；H5页面安全合规检测数据主要是检查H5页面源代码中的<script>标签与标签中嵌套的代码中是否包含诸如XSS漏洞等攻击行为代码，若不含包含攻击性代码，则ad_h5设置为0，否则设置为1。识别完毕后，将四个结果内容，即ad_word、ad_sound、ad_photo、ad_h5，连同广告素材数据和广告主数据发送到区块链中，以便触发智能合约进行广告审核和上架判断。

[0082] 步骤S30：获取各广告审核员的审核指数，并根据所述审核指数对广告审核员进行排序，得到广告审核员数据列表，其中，所述审核指数与智能合约推荐对应广告审核员进行广告审核的推荐程度相关。

[0083] 需要说明的是，人工审核通过区块链智能合约分发多人，人工审核候选名单最初由广告审核平台提供，随着人工审核和用户对广告的反馈，装置会动态调整人工审核候选名单中每个人的建议审核指数。

[0084] 可以理解的是，审核指数与智能合约推荐对应广告审核员进行广告审核的推荐程度相关，审核指数最初为1，随着广告审核员审核质量在用户侧的反馈，则会发生变化，审核指数越低，则智能合约不推荐该广告审核员进行某类型广告的审核。

[0085] 如图4所示，图4为本实施例广告审核方法中的智能合约数据结构设计图，智能合约数据结构包含四个数据区，分别是合约元数据区、合约数据去、合约区以及合约数据校验区，其中，合约元数据区(ContractMetaData)主要包含合约的元数据信息，如合约编号、合约名称、合约版本以及合约触发方式；合约数据区(ContractData)主要存放业务的静态数据，静态数据包括广告编号、广告主编号、广告标题、广告描述、广告类型、广告访问地址、广告主状态、广告主缴费情况、历史广告合规、文字合规检测数据、音频合规检测数据、图像视频合规检测数据、H5页面安全合规检测数据、人工审核员名单数据；合约区(Contract)主要包含若干个合约函数组成的函数列表及业务接口回调，如审核合约函数、上架合约函数、人工审核分发函数合约函数以及回调函数；合约数据校验区(ContractVerifyData)主要包含上述三个数据区，即合约元数据区、合约数据区、合约区数据的校验，以便在区块链中进行验证数据的完整性和真实性，对应为合约元数据校验metaVerify、合约数据区校验contractDataVerify、合约区校验contractVerify。

[0086] 在具体实现中，广告类型(整形字段类型，字段命名ad_type)有多种，通过整数不同数字标注不同广告类型，常见的广告类型有图片型、H5型、PC型、音频型、视频型等。此处图片型时ad_type＝1，H5型ad_type＝2，PC型时ad_type＝3，音频型时ad_type＝4，视频型时ad_type＝5。合约编号(字符串字段类型，字段命名contract_num)、合约名称(字符串字段类型，字段命名contract_name)、合约版本(字符串字段类型，字段命名contract_version)、合约触发方式(整形字段类型，字段命名contract_runway)；广告编号(整形字段类型，字段命名ad_num)、广告主编号(整形字段类型，字段命名ad_owner_num)、广告标题(字符串字段类型，字段命名ad_title)、广告描述(字符串字段类型，字段命名ad_des)、广告类型(整形字段类型，字段命名ad_type)、广告访问地址(字符串字段类型，字段命名ad_url)、广告主状态(整形字段，字段命名ad_owner_status，其中ad_owner_status＝0位默认状态，标识正常；ad_owner_status＝1标识广告主被封禁)、广告主缴费情况(整形字段，字段命名ad_owner_money，其中ad_owner_money为负值时标识广告主已欠费)、历史广告合规(整形字段，字段命名ad_owner_ads，标识历史广告的合规情况，ad_owner_ads＝0标识合规，ad_owner_ads＝1标识不合规)、文字合规检测数据(整形字段，字段命名ad_word，其中ad_word＝0标识广告文字检测正常，ad_word＝1标识文字检测异常)、音频合规检测数据(整形字段，字段命名ad_sound，其中ad_sound＝0标识广告音频检测正常，ad_sound＝1标识音频检测异常)、图像视频合规检测数据(整形字段，字段命名ad_photo，其中ad_photo＝0标识广告图像视频检测正常，ad_photo＝1标识图像视频检测异常)、H5页面安全合规检测数据(整形字段，字段命名ad_h5，其中ad_h5＝0标识广告H5页面安全检测正常，ad_h5＝1标识H5页面安全检测异常)、人工审核员名单数据(JSONArray(数组)类型，字段命名man_checkers，JSONArray(数组)中每个JSON(数据)中的核心字段是审核员的ID、可审核广告类型、邮箱、手机号码)

[0087] 值得说明的是，业务接口回调接口为HTTP回调方式，接口回调的描述包含、接口API的URL、接口的访问方式，如get方式还是POST方式、接口参数列表。合约函数的格式如下所示：

[0088] 返回值function合约名(输入参数列表){

[0089] 函数具体实现

[0090] }

[0091] 其中，返回值为SQL，以便SQL能高效执行。

[0092] 广告审核合约函数、上架合约函数以及人工审核分发函数合约函数的实现代码如下：

[0093] 广告审核合约

[0094] Boolean function adVerify(ContractMetaData contractMetaData，ContractData contractData){

[0095] / / 具体判断广告审核流程

[0096] return TRUE or FALSE；

[0097] }

[0098] 广告上架合约

[0099] Boolean function adonline(Boolean isVerify，ContractData contractData){

[0100] / / 具体判断广告上架

[0101] return TRUE or FALSE；

[0102] }

[0103] 人工审核分发合约

[0104] Boolean function send2mancheckers(Boolean isVerify，ContractDatacontractData){

[0105] / / 具体执行人工审核分发

[0106] return TRUE or FALSE；

[0107] }

[0108] 进一步地，所述获取各广告审核员的审核指数，并根据所述审核指数对广告审核员进行排序，得到广告审核员数据列表，包括：获取广告审核员名单数据，其中，所述广告审核员名单数据中包括广告审核员名单以及各广告审核员可审核的广告类型；根据所述广告类型与所述各广告审核员可审核的广告类型对所述广告审核员名单进行筛选，得到目标广告审核员名单；获取所述目标广告审核员名单中的各广告审核员的审核指数；根据所述目标广告审核员名单中的各广告审核员的审核指数对所述目标广告审核员名单中的广告审核员进行推荐程度排序，得到广告审核员数据列表。

[0109] 需要说明的是，根据传入的广告类型，即字段ad_type来判断是否进入人工审核阶段，若ad_type＝2或ad_type＝4或ad_type＝5，即广告为H5广告或音频广告或视频广告时，则需要人工审核，根据广告类型ad_type从广告审核员名单数据字段man_checkers中获取对应的审核人员列表，即根据字段man_checkers的值也就是JSONArray(数组)中每个JSON(数据)中的核心字段是审核员的可审核广告类型来判断是否与当前广告类型ad_type匹配，若匹配则说明该JSON(数据)对应的审核员可审核该广告，其中，为了式筛选过程高效，广告审核员名单数据man_checkers中的数据通过数据库进行存储，以便根据广告类型ad_type来检索得到对应的广告审核员。

[0110] 如表2所示，表2为广告审核员配置表，表1中包括字段名称、字段类型以及字段含义，例如，字段int的字段类型为整数类型int，字段含义为数据库中默认自增id，将其视为合约编号；字段man_id的字段类型为整数类型int，字段含义为广告审核员编号；字段man_phone的字段类型为varchar，字段含义为广告审核员手机号码；字段man_mail的字段类型为varchar，字段含义为广告审核员邮箱；字段man_type的字段类型为int，字段含义为广告审核员可审核的广告类型，对应的数值同广告类型ad_type字段的值；字段create_time的字段类型为datetime，字段含义为数据录入时间；字段man_send_index的字段类型为double，字段含义为审核员的审核指数，初始化为1。

[0111] 表2

[0112]

[0113] 在具体实现中，通过筛选，可获得广告可审核的多条广告审核员数据，即目标广告审核员名单，并按审核指数man_send_index由大到小排序，得到广告审核员数据列表List。

[0114] 进一步地，所述获取各广告审核员的审核指数之前，还包括：当检测到用户上传广告观看后的反馈数据时，获取用户观看的广告的广告审核结果以及对应广告审核员的初始审核指数；根据所述广告观看后的反馈数据、所述用户观看的广告的广告审核结果以及所述对应广告审核员的初始审核指数进行计算，得到对应广告审核员的当前审核指数；根据所述当前审核指数对所述初始审核指数进行更新，得到各广告审核员的审核指数。

[0115] 在具体实现中，每个审核员的初始化审核指数man_send_index＝1；当有用户反馈数据上传到区块链时，将用户反馈的结果写入广告与审核员的分发和审核结果记录表中的用户对广告观看后的反馈结果user_result字段；根据用户对广告观看后的反馈结果user_result，计算该广告对应的多个审核员的审核指数man_send_index，计算公式为：man_send_index＝(user_result+man_result) / 2×man_send_index / 该审核员审核次数，更新广告审核员配置表中对应审核员的初始审核指数man_send_index。

[0116] 步骤S40：根据所述广告审核员数据列表确定目标广告审核员，并通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果。

[0117] 需要说明的是，对AI识别能力偏弱的如H5页面广告、视频广告继续通过智能合约将相关广告素材发给目标广告审核员进行人工审核，得到人工审核结果。

[0118] 步骤S50：将所述自动化审核结果和所述人工审核结果发布至所述区块链，并通过所述智能合约根据所述自动化审核结果和所述人工审核结果进行计算，得到广告审核结果。

[0119] 需要说明的是，广告审核规则通过AI识别和区块链智能合约进行自动化的审核，同时共识规则根据广告的类型将需要人工审核的素材通过智能合约分发多人审核并将审核结果记录在区块链的共识结果中，自动化审核计日工和人工审核结果经区块链共识并存储上链后则生成广告审核结果，即广告是否审核通过，如果审核通过可以告知广告业务平台进行自动上架，将广告审核和上架的过程自动化、智能化、提高了审核效率，通过智能合约计算选出的多人审核也避免了原来单个审核人员和广告主利益交换，进而保证了合规广告的公平公正的上架和排名。

[0120] 进一步地，所述将所述自动化审核结果和所述人工审核结果发布至所述区块链，并通过所述智能合约根据所述自动化审核结果和所述人工审核结果进行计算，得到广告审核结果之后，还包括：在广告审核结果为审核通过时，通过所述智能合约将所述广告素材的状态变成为已审核状态；检测所述广告主数据，根据所述广告主数据确定所述广告主是否存在欠费行为；若不存在，则根据所述广告主数据确定所述广告主是否存在历史广告违规整改行为；若不存在，则通过所述智能合约对所述广告素材进行自动上架。

[0121] 需要说明的是，审核完成后，对应广告主的某条广告数据has_verify标识为1，显示已审核，以实时触发为例，当审核合约执行结果为TRUE时，可继续执行上架合约，若广告主欠费(即ad_owner_money为负值)或有历史广告有违规整改情况(即ad_owner_ads＝1)上架结果为不上架；否则进行自动上架。当审核合约执行结果为FALSE时，继续执行上架合约，上架结果为不上架；本地节点完成合约执行后，一方面将结果存在本地，即设置广告在本地节点的上架情况local_online的值，另一方面将结果打包成区块，发布到联盟链各个节点中进行共识。当各个节点收到其他结点发来的某条广告审核和上架的结果后，根据共识算法的规则，判断是否同意相关结果，如果同意，则将本地的联盟链共识结果字段进行修改，即广告在区块链各个节点的上架情况chain__online＝1。当完成全链共识和结果存储后，将结果回调给业务方。

[0122] 在具体实现中，将广告审核和上架的结果在区块链中达成共识和存储后，发起结果共识的区块发起对业务层的回调，以便告知业务层结果。接口回调的描述包含接口API的URL(网址)、接口的访问方式，如get方式或POST方式以及接口参数列表。本实施例中采用的是基于消息队列MQ的接口回调，假设URL＝https: / / AA.com / BB / CC，用POST形式访问，则将接口参数作为POST请求中body部分的内容传输，具体实现代码如下：

[0123] {

[0124] "verify_type":1,

[0125] "online_ids":[1111,223,3332]

[0126] }

[0127] 其中，verify_type代表审核数据类型，verify_type＝1时标识广告审核通过，对应的广告ID在online_ids的JSONArray数组中；action_type＝0时标识广告审核不通过，此时online_ids的JSONArray数组为空。

[0128] 广告审核和上架的结果通过网址URL＝https: / / AA.com / BB / CC可以发送到消息队列MQ中，如Kafka，广告主和审核员可以分别在广告配置平台和广告运营平台查看广告的审核结果和上架情况。

[0129] 本实施例通过基于预设广告审核规则和预设上架规则生成智能合约，并将所述智能合约发布至区块链；当接收到广告主发送的广告素材数据以及广告主数据时，基于所述广告素材数据和所述广告主数据进行自动化审核，得到自动化审核结果；获取各广告审核员的审核指数，并根据所述审核指数对广告审核员进行排序，得到广告审核员数据列表，其中，所述审核指数与智能合约推荐对应广告审核员进行广告审核的推荐程度相关；根据所述广告审核员数据列表确定目标广告审核员，并通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果；将所述自动化审核结果和所述人工审核结果发布至所述区块链，并通过所述智能合约根据所述自动化审核结果和所述人工审核结果进行计算，得到广告审核结果。通过上述方式，将智能合约发布至区块链，通过智能合约对广告素材分别进行自动化审核和人工审核，并通过智能合约根据自动化审核结果和人工审核结果进行计算，得到广告审核结果，解决了广告审核效率低且存在审核人员和广告主间人为利益交换的空间和风险的问题，提高审核效率，避免审核人员和广告主利益交换，保证合规广告公平公正的上架和排名。

[0130] 参考图5，图5为本发明广告审核方法第二实施例的流程示意图。

[0131] 基于上述第一实施例，本实施例广告审核方法中所述步骤S40，包括：

[0132] 步骤S401：获取广告审核员数据列表的长度以及单次发送广告素材的数量阈值。

[0133] 需要说明的是，根据广告类型ad_type在获得该广告可审核的多条广告审核员数据，并按审核指数man_send_index由大到小排序，这些数据组成广告审核员数据列表List。

[0134] 可以理解的是，单次发送广告素材的数量阈值即每次有效发送审核人员的数量K；

[0135] 步骤S402：将所述广告审核员数据列表的长度与所述单次发送广告素材的数量阈值进行比较。

[0136] 可以理解的是，若广告审核员数据列表List的长度小于等于K，则将广告审核员数据列表List中的广告审核员均作为目标广告审核员。

[0137] 步骤S403：若所述广告审核员数据列表的长度大于所述单次发送广告素材的数量阈值，则根据所述广告审核员数据列表中各广告审核员的审核指数确定审核阈值。

[0138] 需要说明的是，若广告审核员数据列表List的长度大于K，则根据广告审核员数据列表List中各审核员的审核指数man_send_index确定审核阈值k。

[0139] 步骤S404：根据所述审核阈值对所述广告审核员数据列表中的广告审核员进行分组，得到第一分组和第二分组，其中，所述第一分组中广告审核员的审核指数大于所述第二分组中广告审核员的审核指数。

[0140] 需要说明的是，审核阈值k将广告审核员数据列表List中的审核员分为2组，即审核指数man_send_index大于k的审核员分在分组1，编号List1；审核指数man_send_index≤k的审核员分在分组2，编号List2。

[0141] 值得说明的是，在分组1中，由于审核指数man_send_index大于k，则说明该组中审核的结果通过用户验证广告质量上整体良好，为主推荐的审核员候选名单；反之，由于审核指数man_send_index≤k，则说明该组中审核的结果通过用户验证广告质量上整体一般，为非主推荐的审核员候选名单。

[0142] 步骤S405：计算所述第一分组和所述第二分组中的广告审核员数量，得到第一数量和第二数量，其中，第一数量大于第二数量。

[0143] 需要说明的是，计算分组1和分组2中本轮广告审核中的审核员数量，即第一数量c1、第二数量c2，其中c1+c2＝K，且c1＞c2。

[0144] 进一步地，所述计算所述第一分组和所述第二分组中的广告审核员数量，得到第一数量和第二数量，包括：对所述数量阈值进行拆解，得到若干个计算数对；对所述若干个计算数对进行随机排序，得到计算数对序列；根据所述数量阈值确定第一随机数，其中，所述第一随机数小于所述数量阈值；根据所述第一随机数从所述计算数对序列中获取目标计算数对；根据所述目标计算数对，确定第一数量和第二数量。

[0145] 需要说明的是，将数量阈值K拆解为2个数的相加，得到若干个相加的计算数对，如1+2＝3，则1和2为一组数对，随机对多个数对进行排序，在0到K的范围内生成序号X，则取排序后第X组数对的值(x,y)，则第一数量c1＝x、第二数量c2＝y。

[0146] 步骤S406：根据第一数量和第二数量对所述第一分组和第二分组中的广告审核员进行筛选，得到目标广告审核员。

[0147] 需要说明的是，计算分组1和分组2中对应的c1、c2个审核员。即分别在分组1中选出c1个审核员，在分组2中选出c2个审核员。当分组1的长度小于等于c1时，则分组1全部为有效审核员；不足的审核员从分组2中获取。

[0148] 当分组1的长度大于c1时，则在分组1选取审核指数man_send_index排名为前c1的为有效审核员；分组2的选取与分组1的选取类似，当分组1小于等于c1时，分组2需将对应缺少数量的审核员补足。补足或选取审核员依然是通过审核指数man_send_index排名来获取对应数量的审核员，即整体上让审核结果有效性更好的审核员来审核，以降低广告曝光到用户后的投诉等负面影响。

[0149] 步骤S407：通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果。

[0150] 需要说明的是，广告审核员数据列表List进行审核人员的选择，选出目标广告审核员，按照表1中对应的短信或邮箱，分发需要目标广告审核员审核的广告素材，审核员收到广告素材并将审核结果(通过或不通过)通过接口发给区块链的某节点。

[0151] 如表3所示，表3为广告与审核员的分发和审核结果记录表，表3中包括字段名称、字段类型以及字段含义，例如，字段int的字段类型为int，字段含义为数据库中默认自增id，将其视为合约编号；字段man_id的字段类型为int，字段含义为广告审核员编号；字段ad_num的字段类型为int，字段含义为广告编号；字段man_result的字段类型为int，字段含义为广告审核结果，1标识通过，0标识不通过；字段user_result的字段类型为int，字段含义为用户对广告观看后的反馈结果，1标识通过，0标识不通过；字段create_time的字段类型为datetime，字段含义为数据录入时间。

[0152] 表3

[0153]

[0154]

[0155] 在具体实现中，区块链中某节点接收到多个审核员对某一广告的审核结果，这些审核结果是陆陆续续分发到区块链上，区块链某节点每次均将某个审核员对该广告的审核结果打包到区块中，并发布到区块链，经共识算法在区块链各个节点形成共识并存储。若表2中对应广告编号ad_num的所有广告审核员编号man_id均已获得广告审核结果man_result，则说明人工审核的多个结果目前已经得到，此时即可进入到广告审核的自动合约逻辑。

[0156] 需要说明的是，联盟链区块链底层服务分为区块链联盟链的各个联盟节点管理平台和区块链节点服务，联盟节点管理平台主要包含联盟管理、联盟秘钥生成、加密、平台登录及鉴权、智能合约下发等子模块；区块链节点服务主要包含区块设计和打包、P2P网络建设、区块上链的共识算法、区块的本地化存储等子模块。

[0157] 如图6所示，图6为本实施例中的联盟链区块链底层服务部署架构图，联盟链底层服务先要启动联盟链的联盟超级节点，该节点具备联盟链节点管理能力，该能力通过联盟节点管理平台提供，该平台启动后，联盟链中各个节点均可自行启动，节点启动时会和联盟节点管理平台进行数据交互，主要涉及联盟节点列表获取以便和其他节点建立通信网络、联盟节点加密解密的公私钥获取等。

[0158] 值得说明的是，当合约需要生成时，按智能合约的数据结构填充对应字段的值，即可完成智能合约数据的组装，随后在区块链联盟链的管理后台合约发布上链，进入联盟链中进行共识和各个节点的合约本地化存储。经过区块链各个节点按共识算法实现共识后，需对智能合约进行本地化存储。

[0159] 如表4所示，表4为智能合约在本地关系型数据库的存储表，表4中包括字段名称、字段类型以及字段含义，字段名称包括int、contract_name、contract_version、contract_runway等，字段类型包括int、Varchar(100)、datetime、tinyint等，字段含义包括数据库中默认自增id，将其视为合约编号、合约名称、合约版本、合约触发方式、合约元数据存储、合约数据存储、合约存储、合约元数据校验、合约数据区校验、合约区数据的校验、数据录入时间、合约状态，当被下线时值为1，默认为0，是非下线状态等，为了方便读写和可维护性，表中的字段与合约中的字段命名一致。合约中的元数据、数据区、合约区的数据及对应的校验数据分别写入表中的第1-10个字段。即完成本地数据的存储。

[0160] 表4

[0161]

[0162] 可以理解的是，合约数据校验区(ContractVerifyData)对合约元数据区、合约数据区、合约区数据的数据进行校验，以便在区块链中进行验证数据的完整性和真实性，计合约元数据校验metaVerify、合约数据区校验contractDataVerify、合约区数据的校验contractVerify。校验的数据根据各组数据的JSON(数据)的MD5值计算得到。以元数据区的数据验证数据生成为例：将合约编号(字符串字段类型，字段命名contract_num)、合约名称(字符串字段类型，字段命名contract_name)、合约版本(字符串字段类型，字段命名contract_version)、合约触发方式(整形字段类型，字段命名contract_runway)与其对应的值组织成JSON格式，即：

[0163] {

[0164] "keys":["contract_num","contract_name","contract_version","contract_runway"],

[0165] }

[0166] 将JSON按照MD5算法生成合约元数据校验metaVerify，即合约元数据校验metaVerify＝MD5(JSON)。

[0167] 在具体实现中，为了方便区块链对数据的解析，广告素材数据和广告主数据、AI检测数据经打包进入区块后，在联盟链中形成共识，即可进行本地化存储，进行根据数据的情况触发智能合约的执行。在联盟链共识需要校验数据的有效性，合约数据校验区(ContractVerifyData)对合约数据区、合约区数据的校验，以便在区块链中进行验证数据的完整性和真实性包括：将合约区数据(含静态数据区和动态数据区)的各个字段提取并组织成JSON格式，即：

[0168] {

[0169] "keys":["contract_num","contract_name","contract_version","contract_runway"],

[0170] }

[0171] JSON按照MD5算法生成合约数据校验contractDataVerify1，即合约元数据校验contractDataVerify1＝MD5(JSON)；通过合约元数据区(ContractMetaData)的合约编号(字符串字段类型，字段命名contract_num)，查询对应合约的contractDataVerify；若contractDataVerify1＝contractDataVerify，则说明数据有效，继续进行共识；若contractDataVerify1≠contractDataVerify，则数据无效，终止校验和共识。

[0172] 如表5所示，表5为广告在联盟链节点中的存储表，表5中包括字段名称、字段类型以及字段含义，字段名称包括int、ad_num、ad_owner_num等，字段类型包括int、Varchar(100)、Varchar(500)等、字段含义包括数据库中默认自增id，将其视为合约编号、广告编号、广告主编号、广告访问地址等。

[0173] 表5

[0174]

[0175]

[0176]

[0177] 可以理解的是，当广告主有广告被下架或要求整改时，ad_owner_ads设置为1。当广告素材数据和广告主数据、AI检测数据存储完成后，即可根据对应的协议合约规定的合约触发方式(整形字段类型，字段命名contract_runway)来进行合约触发。常见有实时触发和定时触发，实时触发时，contract_runway＝0；定时触发时，将触发时间以整数串形式编写，如2023-04-10 10:00:00可以写成20230410100000，即contract_runway＝20230410100000。

[0178] 值得说明的是，广告审核合约的内容具体包括：进行广告主状态检查，即检查合约中的ad_owner_status字段，若ad_owner_status＝0位默认状态标识广告主正常，ad_owner_status＝1标识广告主被封禁。当广告主被封禁时，广告审核合约返回FALSE，广告审核合约执行完成；当广告主未被封禁时，进行广告素材完备性检查，即检查合约中的广告编号ad_num、广告主编号ad_owner_num、广告标题ad_title、广告描述ad_des、广告类型ad_type、广告访问地址ad_url字段是否均有数值，当其中之一没有数据时，合约返回FALSE，进行多人人工审核结果检测；若均有数据，则进行AI检查结果审核，即检查合约中的文字合规检测数据ad_word、音频合规检测数据ad_sound、图像视频合规检测数据ad_photo、H5页面安全合规检测数据ad_h5。当ad_word＝1或ad_sound＝1或ad_photo＝1或ad_h5＝1，合约返回FALSE，进行多人人工审核结果检测，就若某广告编号ad_num，超过半数的广告审核员man_id的广告审核结果man_result为通过，则人工审核结果判定为通过，返回TRUE，否则返回FALSE，合约执行完成。

[0179] 上架合约的内容具体包括：根据传入参数isVerify来判断，若isVerify＝FALSE，则说明广告审核不通过，上架合约返回FALSE，上架合约执行完成；若广告审核通过，即isVerify＝TRUE，判断合约数据区中的广告主欠费情况，若合约数据区中ad_owner_mone为负值，即广告主已欠费，则此时合约返回FALSE，合约执行完成；若合约数据区中ad_owner_mone为正值，说明广告主不欠费，广告审核通过且广告主不欠费，根据广告主ID，查看广告主是否有历史广告有违规整改情况，即ad_owner_ads字段，若ad_owner_ads＝1则说明有待整改的广告，此时合约返回FALSE，否则合约返回TRUE，合约执行完成。

[0180] 本实施例通过获取广告审核员数据列表的长度以及单次发送广告素材的数量阈值；将所述广告审核员数据列表的长度与所述单次发送广告素材的数量阈值进行比较；若所述广告审核员数据列表的长度大于所述单次发送广告素材的数量阈值，则根据所述广告审核员数据列表中各广告审核员的审核指数确定审核阈值；根据所述审核阈值对所述广告审核员数据列表中的广告审核员进行分组，得到第一分组和第二分组，其中，所述第一分组中广告审核员的审核指数大于所述第二分组中广告审核员的审核指数；计算所述第一分组和所述第二分组中的广告审核员数量，得到第一数量和第二数量，其中，第一数量大于第二数量；根据第一数量和第二数量对所述第一分组和第二分组中的广告审核员进行筛选，得到目标广告审核员；通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果，提高了人工审核结果的准确率。

[0181] 参照图7，图7为本发明广告审核装置第一实施例的结构框图。

[0182] 如图7所示，本发明实施例提出的广告审核装置包括：

[0183] 发布模块10，用于基于预设广告审核规则和预设上架规则生成智能合约，并将所述智能合约发布至区块链。

[0184] 审核模块20，用于当接收到广告主发送的广告素材数据以及广告主数据时，基于所述广告素材数据和所述广告主数据进行自动化审核，得到自动化审核结果。

[0185] 排序模块30，用于获取各广告审核员的审核指数，并根据所述审核指数对广告审核员进行排序，得到广告审核员数据列表，其中，所述审核指数与智能合约推荐对应广告审核员进行广告审核的推荐程度相关。

[0186] 所述审核模块20，还用于根据所述广告审核员数据列表确定目标广告审核员，并通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果。

[0187] 确定模块40，用于将所述自动化审核结果和所述人工审核结果发布至所述区块链，并通过所述智能合约根据所述自动化审核结果和所述人工审核结果进行计算，得到广告审核结果。

[0188] 本实施例通过基于预设广告审核规则和预设上架规则生成智能合约，并将所述智能合约发布至区块链；当接收到广告主发送的广告素材数据以及广告主数据时，基于所述广告素材数据和所述广告主数据进行自动化审核，得到自动化审核结果；获取各广告审核员的审核指数，并根据所述审核指数对广告审核员进行排序，得到广告审核员数据列表，其中，所述审核指数与智能合约推荐对应广告审核员进行广告审核的推荐程度相关；根据所述广告审核员数据列表确定目标广告审核员，并通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果；将所述自动化审核结果和所述人工审核结果发布至所述区块链，并通过所述智能合约根据所述自动化审核结果和所述人工审核结果进行计算，得到广告审核结果。通过上述方式，将智能合约发布至区块链，通过智能合约对广告素材分别进行自动化审核和人工审核，并通过智能合约根据自动化审核结果和人工审核结果进行计算，得到广告审核结果，解决了广告审核效率低且存在审核人员和广告主间人为利益交换的空间和风险的问题，提高审核效率，避免审核人员和广告主利益交换，保证合规广告公平公正的上架和排名。

[0189] 在一实施例中，所述排序模块30，还用于获取广告审核员名单数据，其中，所述广告审核员名单数据中包括广告审核员名单以及各广告审核员可审核的广告类型；根据所述广告类型与所述各广告审核员可审核的广告类型对所述广告审核员名单进行筛选，得到目标广告审核员名单；获取所述目标广告审核员名单中的各广告审核员的审核指数；根据所述目标广告审核员名单中的各广告审核员的审核指数对所述目标广告审核员名单中的广告审核员进行推荐程度排序，得到广告审核员数据列表。

[0190] 在一实施例中，所述审核模块20，还用于获取广告审核员数据列表的长度以及单次发送广告素材的数量阈值；将所述广告审核员数据列表的长度与所述单次发送广告素材的数量阈值进行比较；若所述广告审核员数据列表的长度大于所述单次发送广告素材的数量阈值，则根据所述广告审核员数据列表中各广告审核员的审核指数确定审核阈值；根据所述审核阈值对所述广告审核员数据列表中的广告审核员进行分组，得到第一分组和第二分组，其中，所述第一分组中广告审核员的审核指数大于所述第二分组中广告审核员的审核指数；计算所述第一分组和所述第二分组中的广告审核员数量，得到第一数量和第二数量，其中，第一数量大于第二数量；根据第一数量和第二数量对所述第一分组和第二分组中的广告审核员进行筛选，得到目标广告审核员；通过所述智能合约将广告素材分配给所述目标广告审核员进行人工审核，得到人工审核结果。

[0191] 在一实施例中，所述审核模块20，还用于对所述数量阈值进行拆解，得到若干个计算数对；对所述若干个计算数对进行随机排序，得到计算数对序列；根据所述数量阈值确定第一随机数，其中，所述第一随机数小于所述数量阈值；根据所述第一随机数从所述计算数对序列中获取目标计算数对；根据所述目标计算数对，确定第一数量和第二数量。

[0192] 在一实施例中，所述排序模块30，还用于当检测到用户上传广告观看后的反馈数据时，获取用户观看的广告的广告审核结果以及对应广告审核员的初始审核指数；根据所述广告观看后的反馈数据、所述用户观看的广告的广告审核结果以及所述对应广告审核员的初始审核指数进行计算，得到对应广告审核员的当前审核指数；根据所述当前审核指数对所述初始审核指数进行更新，得到各广告审核员的审核指数。

[0193] 在一实施例中，所述审核模块20，还用于根据所述广告素材数据确定广告类型；根据所述广告类型对广告素材进行合规性检测，得到检测结果，所述合规性检测至少包括文字合规检测、音频合规检测、图像视频合规检测以及H5页面安全合规检测中的至少一项；基于所述广告素材数据、所述广告主数据以及所述检测结果进行自动化审核，得到自动化审核结果。

[0194] 在一实施例中，所述确定模块40，还用于在广告审核结果为审核通过时，通过所述智能合约将所述广告素材的状态变成为已审核状态；检测所述广告主数据，根据所述广告主数据确定所述广告主是否存在欠费行为；若不存在，则根据所述广告主数据确定所述广告主是否存在历史广告违规整改行为；若不存在，则通过所述智能合约对所述广告素材进行自动上架。

[0195] 此外，为实现上述目的，本发明还提出一种广告审核设备，所述广告审核设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的广告审核程序，所述广告审核程序配置为实现如上文所述的广告审核方法的步骤。

[0196] 由于本广告审核设备采用了上述所有实施例的全部技术方案，因此至少具有上述实施例的技术方案所带来的所有有益效果，在此不再一一赘述。

[0197] 此外，本发明实施例还提出一种存储介质，所述存储介质上存储有广告审核程序，所述广告审核程序被处理器执行时实现如上文所述的广告审核方法的步骤。

[0198] 由于本存储介质采用了上述所有实施例的全部技术方案，因此至少具有上述实施例的技术方案所带来的所有有益效果，在此不再一一赘述。

[0199] 应当理解的是，以上仅为举例说明，对本发明的技术方案并不构成任何限定，在具体应用中，本领域的技术人员可以根据需要进行设置，本发明对此不做限制。

[0200] 需要说明的是，以上所描述的工作流程仅仅是示意性的，并不对本发明的保护范围构成限定，在实际应用中，本领域的技术人员可以根据实际的需要选择其中的部分或者全部来实现本实施例方案的目的，此处不做限制。

[0201] 另外，未在本实施例中详尽描述的技术细节，可参见本发明任意实施例所提供的广告审核方法，此处不再赘述。

[0202] 此外，需要说明的是，在本文中，术语"包括”、"包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者系统不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者系统所固有的要素。在没有更多限制的情况下，由语句"包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者系统中还存在另外的相同要素。

[0203] 应该理解的是，虽然本申请实施例中的流程图中的各个步骤按照箭头的指示依次显示，但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明，这些步骤的执行并没有严格的顺序限制，其可以以其他的顺序执行。而且，图中的至少一部分步骤可以包括多个子步骤或者多个阶段，这些子步骤或者阶段并不必然是在同一时刻执行完成，而是可以在不同的时刻执行，其执行顺序也不必然是依次进行，而是可以与其他步骤或者其他步骤的子步骤或者阶段的至少一部分轮流或者交替地执行。

[0204] 上述本发明实施例序号仅仅为了描述，不代表实施例的优劣。

[0205] 通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如只读存储器(Read Only Memory，ROM) / RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，或者网络设备等)执行本发明各个实施例所述的方法。

[0206] 以上仅为本发明的优选实施例，并非因此限制本发明的专利范围，凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本发明的专利保护范围内。< / script>

Claims

1. An advertisement review method, characterized in that: The method comprises: Generate a smart contract based on preset advertising review rules and preset listing rules, and publish the smart contract to the blockchain; When receiving the advertising material data and the advertiser data sent by the advertiser, performing an automated review based on the advertising material data and the advertiser data to obtain an automated review result; Obtain the review index of each advertising reviewer and sort the advertising reviewers according to the review index to obtain a list of advertising reviewer data, wherein the review index is related to the degree to which the smart contract recommends the corresponding advertising reviewer to conduct advertising review; Determine a target advertising reviewer based on the advertising reviewer data list, and assign the advertising material to the target advertising reviewer for manual review through the smart contract to obtain a manual review result; Publishing the automated review result and the manual review result to the blockchain, and performing calculations based on the automated review result and the manual review result through the smart contract to obtain an advertisement review result; Before obtaining the audit index of each advertisement auditor, the method further includes: When feedback data after a user uploads an advertisement is detected, the advertisement review result of the advertisement viewed by the user and the initial review index of the corresponding advertisement reviewer are obtained, where the initial review index is 1; Calculate the current review index of the corresponding advertising reviewer based on the feedback data after viewing the advertisement, the advertisement review result of the advertisement viewed by the user, and the initial review index of the corresponding advertising reviewer; The initial review index is updated according to the current review index to obtain the review index of each advertising reviewer.

2. The method according to claim 1, wherein The step of obtaining the audit index of each advertising auditor and sorting the advertising auditors according to the audit index to obtain an advertising auditor data list includes: Acquire advertising reviewer list data, wherein the advertising reviewer list data includes a list of advertising reviewers and the types of advertisements that each advertising reviewer can review; Screening the list of advertising reviewers according to the advertisement type and the advertisement type that each advertisement reviewer can review, to obtain a target advertisement reviewer list; Obtaining the review index of each advertisement reviewer in the target advertisement reviewer list; The advertising reviewers in the target advertising reviewer list are ranked by recommendation according to the review index of each advertising reviewer in the target advertising reviewer list to obtain an advertising reviewer data list.

3. The method according to claim 1, wherein The step of determining a target advertising reviewer according to the advertising reviewer data list, and assigning the advertising material to the target advertising reviewer for manual review through the smart contract, and obtaining a manual review result includes: Get the length of the ad reviewer data list and the threshold for the number of ad creatives sent at a time; Comparing the length of the advertisement auditor data list with the threshold value of the number of advertisement materials sent at a single time; If the length of the advertising reviewer data list is greater than the threshold value of the number of advertising materials sent at a single time, determining the review threshold value according to the review index of each advertising reviewer in the advertising reviewer data list; Grouping the advertising auditors in the advertising auditor data list according to the audit threshold to obtain a first group and a second group, wherein the audit index of the advertising auditors in the first group is greater than the audit index of the advertising auditors in the second group; Calculating the number of advertising reviewers in the first group and the second group to obtain a first number and a second number, wherein the first number is greater than the second number; Screening the advertising reviewers in the first group and the second group according to the first quantity and the second quantity to obtain target advertising reviewers; The advertising material is assigned to the target advertising reviewer through the smart contract for manual review to obtain a manual review result.

4. The method according to claim 3, wherein The calculating the number of advertising reviewers in the first group and the second group to obtain the first number and the second number includes: Decomposing the quantity threshold to obtain a plurality of calculation number pairs; Randomly sorting the plurality of calculated number pairs to obtain a calculated number pair sequence; determining a first random number according to the quantity threshold, wherein the first random number is smaller than the quantity threshold; Obtaining a target calculation number pair from the calculation number pair sequence according to the first random number; Calculate number pairs according to the target to determine a first quantity and a second quantity.

5. The method according to claim 1, wherein The automated review based on the advertising material data and the advertiser data to obtain the automated review result includes: Determine the advertisement type according to the advertisement creative data; Performing compliance testing on the advertising material according to the advertisement type to obtain a test result, wherein the compliance testing includes at least one of text compliance testing, audio compliance testing, image and video compliance testing, and H5 page security compliance testing; An automated review is performed based on the advertising material data, the advertiser data, and the detection result to obtain an automated review result.

6. The method according to any one of claims 1 to 5, characterized in that The automated review result and the manual review result are published to the blockchain, and after the smart contract is used to calculate based on the automated review result and the manual review result to obtain the advertisement review result, the method further includes: When the advertisement review result is approved, the status of the advertisement material is changed to the reviewed status through the smart contract; detecting the advertiser data, and determining whether the advertiser has any outstanding payment behavior based on the advertiser data; If not, determining whether the advertiser has any historical advertising violation rectification behavior based on the advertiser data; If it does not exist, the advertising material is automatically put on the shelf through the smart contract.

7. An automatic advertisement review device, characterized in that: The automatic advertisement review device comprises: A publishing module, configured to generate a smart contract based on preset advertising review rules and preset listing rules, and publish the smart contract to the blockchain; An audit module is configured to, upon receiving advertising material data and advertiser data sent by an advertiser, perform an automated audit based on the advertising material data and the advertiser data to obtain an automated audit result; A sorting module is used to obtain the review index of each advertising reviewer and sort the advertising reviewers according to the review index to obtain a list of advertising reviewer data, wherein the review index is related to the degree of recommendation of the smart contract for the corresponding advertising reviewer to conduct advertising review; The review module is further configured to determine a target advertising reviewer based on the advertising reviewer data list, and assign the advertising material to the target advertising reviewer through the smart contract for manual review to obtain a manual review result; a determination module, configured to publish the automated review result and the manual review result to the blockchain, and calculate, through the smart contract, based on the automated review result and the manual review result to obtain an advertisement review result; The sorting module is also used to obtain the advertising review results of the advertisements viewed by the user and the initial review index of the corresponding advertising reviewer when feedback data after viewing the advertisements uploaded by the user is detected, and the initial review index is 1; calculate based on the feedback data after viewing the advertisements, the advertising review results of the advertisements viewed by the user, and the initial review index of the corresponding advertising reviewer to obtain the current review index of the corresponding advertising reviewer; update the initial review index based on the current review index to obtain the review index of each advertising reviewer.

8. An automatic advertising review device, characterized in that: The automatic advertisement review device includes: a memory, a processor, and an automatic advertisement review program stored in the memory and executable on the processor, wherein the automatic advertisement review program is configured to implement the automatic advertisement review method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores an automatic advertisement review program, which, when executed by a processor, implements the automatic advertisement review method according to any one of claims 1 to 6.

Images